A heart attack is an urgent medical situation. It usually occurs when a blood clot obstructs the heart’s blood supply. Without blood, tissue dehydrates and deteriorates. Treatment may involve medications, medical procedures such as angioplasty or stent placement to open the blocked artery, and lifestyle changes to reduce the risk of future heart problems.

In this article, I assume the role of a Healthcare Data Analyst and carry out exploratory data analysis of Heart Attack dataset of a hospital. From this analysis, insights on the causes of heart diseases is gotten. The dataset used in this analysis was gotten from Kaggle, and it contains information such as: age, sex, resting blood pressure, exercise induced angina, number of major vessels, chest pain type, cholesterol, fasting blood sugar, resting electrocardiographic result, maximum heart rate, ST depression, thallium stress test, slope peak exercise and output.

I used MS SQL Server to clean the dataset and carry out the data exploration on the dataset. This article is divided into several stages: data cleaning, data exploration, insights and recommendation stage.

To gain insights from a dataset, it is important to ensure the dataset is clean enough for doing that. The data cleaning process involves handling missing values, correcting inconsistencies, and formatting the data for analysis. In this project, I ensured the dataset was thoroughly cleaned to ensure the accuracy and integrity of our findings.

Checking if there are duplicate rows

SELECT age, sex, cp, trtbps, chol, fbs, restecg, thalachh, exng, oldpeak, slp, caa, thall, output FROM PersonalProject..HeartAttack GROUP BY age, sex, cp, trtbps, chol, fbs, restecg, thalachh, exng, oldpeak, slp, caa, thall, output HAVING COUNT(*) > 1;

The duplicate row has to be removed to ensure accuracy-- Dropping the duplicate row WITH CTE AS ( SELECT age, sex, cp, trtbps, chol, fbs, restecg, thalachh, exng, oldpeak, slp, caa, thall, output, ROW_NUMBER() OVER (PARTITION BY age, sex, cp, trtbps, chol, fbs, restecg, thalachh, exng, oldpeak, slp, caa, thall, output ORDER BY (SELECT 0)) AS RowNumber FROM PersonalProject..HeartAttack ) DELETE FROM CTE WHERE RowNumber > 1;

Checking for null values

SELECT * FROM PersonalProject..HeartAttack WHERE age IS NULL OR sex IS NULL OR cp IS NULL OR trtbps IS NULL OR chol IS NULL OR fbs IS NULL OR restecg IS NULL OR thalachh IS NULL OR exng IS NULL OR oldpeak IS NULL OR slp IS NULL OR caa IS NULL OR thall IS NULL OR output IS NULL;

Next step in the data cleaning stage is renaming of columns for better readability and understanding.

The columns to be renamed are : cp, trtbps, chol, fbs, restecg, thalachh, exng, oldpeak, slp, caa, thall

BEGIN TRANSACTION; EXEC sp_rename 'PersonalProject..HeartAttack.cp', 'chest_pain_type', 'COLUMN'; EXEC sp_rename 'PersonalProject..HeartAttack.trtbps', 'resting_blood_pressure', 'COLUMN'; EXEC sp_rename 'PersonalProject..HeartAttack.chol', 'cholestoral', 'COLUMN'; EXEC sp_rename 'PersonalProject..HeartAttack.fbs', 'fasting_blood_sugar', 'COLUMN'; EXEC sp_rename 'PersonalProject..HeartAttack.restecg', 'resting_electrocardiographic', 'COLUMN'; EXEC sp_rename 'PersonalProject..HeartAttack.thalachh', 'maximum_heart_rate', 'COLUMN'; EXEC sp_rename 'PersonalProject..HeartAttack.exng', 'exercise_induced_angina', 'COLUMN'; EXEC sp_rename 'PersonalProject..HeartAttack.oldpeak', 'ST_depression', 'COLUMN'; EXEC sp_rename 'PersonalProject..HeartAttack.slp', 'slope_peak_exercise', 'COLUMN'; EXEC sp_rename 'PersonalProject..HeartAttack.caa', 'number_of_major_vessels', 'COLUMN'; EXEC sp_rename 'PersonalProject..HeartAttack.thall', 'thalium_stress_test', 'COLUMN'; COMMIT;

Viewing renamed columns:

SELECT * FROM PersonalProject..HeartAttack

a section of renamed columns

Adding feature values to the columns. This will aid better understanding of the dataset. For the other columns where feature values will be added, new columns has to be created for them and then the feature values added

--Adding feature values for sex ALTER TABLE PersonalProject..HeartAttack ADD sex_new VARCHAR(50); UPDATE PersonalProject..HeartAttack SET sex_new = CASE WHEN sex = 0 THEN 'Female' WHEN sex = 1 THEN 'Male' END; -- Dropping the old column ALTER TABLE PersonalProject..HeartAttack DROP COLUMN sex; -- Rename the new column EXEC sp_rename 'PersonalProject..HeartAttack.sex_new', 'sex', 'COLUMN'; --Adding feature values for chest_pain_type ALTER TABLE PersonalProject..HeartAttack ADD chest_pain_type_new VARCHAR(50); UPDATE PersonalProject..HeartAttack SET chest_pain_type_new = CASE WHEN chest_pain_type = 0 THEN 'typical angina' WHEN chest_pain_type = 1 THEN 'atypical angina' WHEN chest_pain_type = 2 THEN 'non-anginal pain' WHEN chest_pain_type = 3 THEN 'asymptomatic' END; -- Dropping the old column ALTER TABLE PersonalProject..HeartAttack DROP COLUMN chest_pain_type; -- Rename the new column EXEC sp_rename 'PersonalProject..HeartAttack.chest_pain_type_new', 'chest_pain_type', 'COLUMN'; -- Adding feature values for fasting_blood_sugar ALTER TABLE PersonalProject..HeartAttack ADD fasting_blood_sugar_new VARCHAR(50); UPDATE PersonalProject..HeartAttack SET fasting_blood_sugar_new = CASE WHEN fasting_blood_sugar = 0 THEN '<120 mg/dl' WHEN fasting_blood_sugar = 1 THEN '>120 mg/dl' END; -- Dropping the old column ALTER TABLE PersonalProject..HeartAttack DROP COLUMN fasting_blood_sugar; -- Rename the new column EXEC sp_rename 'PersonalProject..HeartAttack.fasting_blood_sugar_new', 'fasting_blood_sugar', 'COLUMN'; -- Adding feature values for resting_electrocardiographic ALTER TABLE PersonalProject..HeartAttack ADD resting_electrocardiographic_new VARCHAR(50); UPDATE PersonalProject..HeartAttack SET resting_electrocardiographic_new = CASE WHEN resting_electrocardiographic = 0 THEN 'normal' WHEN resting_electrocardiographic = 1 THEN 'ST-T wave abnormality' WHEN resting_electrocardiographic = 2 THEN 'ventricular hypertrophy' END; -- Dropping the old column ALTER TABLE PersonalProject..HeartAttack DROP COLUMN resting_electrocardiographic; -- Rename the new column EXEC sp_rename 'PersonalProject..HeartAttack.resting_electrocardiographic_new', 'resting_electrocardiographic', 'COLUMN'; -- Adding feature values for exercise_induced_angina ALTER TABLE PersonalProject..HeartAttack ADD exercise_induced_angina_new VARCHAR(50); UPDATE PersonalProject..HeartAttack SET exercise_induced_angina_new = CASE WHEN exercise_induced_angina = 0 THEN 'no' WHEN exercise_induced_angina = 1 THEN 'yes' END; -- Dropping the old column ALTER TABLE PersonalProject..HeartAttack DROP COLUMN exercise_induced_angina; -- Rename the new column EXEC sp_rename 'PersonalProject..HeartAttack.exercise_induced_angina_new', 'exercise_induced_angina', 'COLUMN'; -- Adding feature values for slope_peak_exercise ALTER TABLE PersonalProject..HeartAttack ADD slope_peak_exercise_new VARCHAR(50); UPDATE PersonalProject..HeartAttack SET slope_peak_exercise_new = CASE WHEN slope_peak_exercise = 0 THEN 'upsloping' WHEN slope_peak_exercise = 1 THEN 'flat' WHEN slope_peak_exercise = 2 THEN 'downsloping' END; -- Dropping the old column ALTER TABLE PersonalProject..HeartAttack DROP COLUMN slope_peak_exercise; -- Rename the new column EXEC sp_rename 'PersonalProject..HeartAttack.slope_peak_exercise_new', 'slope_peak_exercise', 'COLUMN'; -- Adding feature values for thalium_stress_test ALTER TABLE PersonalProject..HeartAttack ADD thalium_stress_test_new VARCHAR(50); UPDATE PersonalProject..HeartAttack SET thalium_stress_test_new = CASE WHEN thalium_stress_test = 0 THEN 'normal 0' WHEN thalium_stress_test = 1 THEN 'normal 1' WHEN thalium_stress_test = 2 THEN 'fixed defect' WHEN thalium_stress_test = 3 THEN 'reversible defect' END; -- Dropping the old column ALTER TABLE PersonalProject..HeartAttack DROP COLUMN thalium_stress_test; -- Rename the new column EXEC sp_rename 'PersonalProject..HeartAttack.thalium_stress_test_new', 'thalium_stress_test', 'COLUMN'; -- Adding feature values for output ALTER TABLE PersonalProject..HeartAttack ADD output_new VARCHAR(50); UPDATE PersonalProject..HeartAttack SET output_new = CASE WHEN output = 0 THEN 'no disease' WHEN output = 1 THEN 'disease' END; -- Dropping the old column ALTER TABLE PersonalProject..HeartAttack DROP COLUMN output; -- Rename the new column EXEC sp_rename 'PersonalProject..HeartAttack.output_new', 'output', 'COLUMN';

I noticed that the name I gave to some columns were wrong, so I had to rename it to the right name.

EXEC sp_rename 'PersonalProject..HeartAttack.cholestoral', 'cholesterol', 'COLUMN' EXEC sp_rename 'PersonalProject..HeartAttack.thalium_stress_test', 'thallium_stress_test', 'COLUMN'

Answering business questions.

What is the average age of patients?

SELECT CAST(AVG(age) AS INTEGER) as AverageAge FROM PersonalProject..HeartAttack

The average age of patients in this dataset is 54.

What is the number of male and female patients?

SELECT sex, COUNT(*) AS TotalNumberOfPatients FROM PersonalProject..HeartAttack GROUP BY sex

The number of male and female patients is 206 and 96, respectively. This shows that more male patients were recorded than females for this analysis.

What is the result of the thallium_stress_test for each sex?

SELECT sex, thallium_stress_test, COUNT(thallium_stress_test) AS TotalThalliumStressTest FROM PersonalProject..HeartAttack GROUP BY sex, thallium_stress_test ORDER BY sex

The result suggests that there are more male patients that have reduced blood flow during exercise (and at rest) compared to the female patients.

Which sex had the most patients with heart diseases?

SELECT sex, COUNT(*) AS CountDisease FROM PersonalProject..HeartAttack WHERE output = 'disease' GROUP BY sex ORDER BY CountDisease DESC;

From the results, it shows that there are more male patients with heart diseases than there are female patients.

Which sex has the highest cholesterol level?

SELECT sex, CAST(AVG(cholesterol) AS INTEGER) AS AverageCholesterol FROM PersonalProject..HeartAttack GROUP BY sex ORDER BY AverageCholesterol DESC;

Female patients were found to have the highest cholesterol level among both sexes.

What Chest pain type resulted more in no heart disease?

SELECT chest_pain_type, COUNT(*) AS CountofNoDisease FROM PersonalProject..HeartAttack WHERE output = 'no disease' GROUP BY chest_pain_type

Amongst the four types of chest pain the patients had, typical angina chest pain mostly resulted to no heart disease.

Which thallium stress test led to higher incidence of heart disease?

SELECT thallium_stress_test, COUNT(*) AS DiseaseCount FROM PersonalProject..HeartAttack WHERE output = 'disease' GROUP BY thallium_stress_test ORDER BY DiseaseCount DESC

Patients with fixed defect were diagnosed with a heart disease.

Is there a relationship between the number of major vessels healthy and the occurrence of heart disease?

SELECT number_of_major_vessels, COUNT(*) AS DiseaseCount FROM PersonalProject..HeartAttack WHERE output = 'disease' GROUP BY number_of_major_vessels ORDER BY number_of_major_vessels;

There is a negative correlation between number of major vessels healthy and the occurrence of heart disease. The higher the number of healthy major vessels, the lesser the occurrence of heart disease.

The average age of patients is 54

The number of male and female patients are 206 and 96, respectively.

There are more male patients that have reduced blood flow during exercise (and at rest) compared to the females.

There are more male patients with heart disease than there are female patients.

Female patients have higher cholesterol level compared to the male patients.

Most patients that had the typical angina chest pain did not have an heart disease.

Fixed defect — a reduced or absent thallium both during exercise and at rest- resulted to the most incidence of heart diseases

The higher the number of healthy major vessels, the lesser the occurrence of heart disease and vice-versa

Patients with lower or no healthy major vessels were more likely to have a heart disease.

Females should be advice to reduce their cholesterol consumption to no more than 300mg per day. They should also be adviced to focus on maintaining a heart-healthy diet that includes a variety of fruits, vegetables, whole grains, lean proteins, and healthy fats.

Medications, improved lifestyle, revascularization — which is done to restore normal blood flow to affected areas of the body and cardiac rehabilitation programs that involve exercise, education, and support can help improve heart health and manage the condition.

More attention should be placed on the male patients to help improve their heart health and prevent occurences that could lead to heart diseases/attack.

By implementing these recommendations, the hospital can help improve the heart health of her patients and manage the condition of patients already diagnosed with a heart disease.

Thank you for reading this article.